Thesis

My PhD thesis was submitted on the 3rd of November 2003 and successfully defended on the 26th of January 2004. The scientific evaluation committee consisted of:

Knowledge Diffusion and Regional Clusters: Lessons from the Danish ICT Industry, 250pp.

Table of Contents

All chapters, except the Introduction and Conclusion, have been published.

Chapter 1: Introduction

  • Dominant schools of geographical agglomeration and regional clustering
  • Theories of entrepreneurship and organizational inheritance
  • Clusters, agglomeration and concentrations: How are they defined?
  • Research questions
  • Structure of the thesis

Chapter 2: On the Danish ICT Sector - Published (2006, edited volume on Routledge)

  • Structure and performance in an international perspective
  • What is behind the relative weakness
  • ICT consumption pattern
  • Regionally clustered strong niches

Chapter 3: Knowledge Diffusion through Informal Contacts: Myth or Reality - Published (2004, Research Policy)

  • Knowledge diffusion and informal contacts
  • Testable propositions
  • Wireless communication cluster in Northern Denmark
  • Importance of informal contacts

Chapter 4: New Firm Formation and Inherited Organizational Capabilities - Published (2010, edited volume on Edward Elgar)

  • Entry and pre-entry organizational capabilities
  • Role of new firm entry in the history of the wireless communication cluster in Northern Denmark

Chapter 5: An Exploration of Regional Labor Mobility - Developed in to several new papers - Published (2010, Social Forces)

  • Labor mobility and clusters
  • Sociological factors and labor mobility

Chapter 6: Geographical Differences in Firm Growth - Published (2004, International Review of Applied Economics)

  • Firm growth and factors of dependence

Chapter 7: Role of Universities in Regional Development - Published (2006, edited volume on Copenhagen Business School Press)

  • Knowledge diffusion and the role of local universities
  • Case-study: Three universities in outer regions in Denmark, Sweden and Finland
  • University-industry interactions

Chapter 8: Conclusion and Policy Perspectives

  • Future research implications

Summary in English

Chapter 1 of the present thesis presents the main research questions, which are studied in the thesis. This is based on a review of the dominant theories of geographical clustering. The central concept of clusters is also discussed.

The point of departure of Chapter 2 is the general worry about the rather weak European performance in the ICT sector. The chapter concentrates on some salient features of the Danish ICT sector in an international context. The main focus is that Denmark on the one hand belongs to the most advanced user nations, measured by conventional indicators for user penetration (such as the amount of PCs, mobile phones and Internet accesses per capita). It is, however, on the other hand rather obvious that the country is not among the major players in the international ICT markets, even when account is taken for the size of the country. The nearby, and in many ways similar, countries of Sweden, Finland and the Netherlands are the living proofs that small countries may play a significantly more visible role in this field. The strengths and weaknesses of the ICT sector in Denmark are analyzed in order to search for possible explanations of this apparent lack of match between the advanced demand and the rather weak supply side of the industry.

The second part of the thesis includes two chapters. Each of these investigates different factors of knowledge diffusion, which are claimed to play a vital role in the emergence of regional clusters. The most commonly mentioned factor of agglomeration and source of knowledge diffusion is informal knowledge flows. This is investigated in Chapter 3. The role of informal networks in the development of regional clusters has received much attention in the literature as described in the review above. Informal contact between employees in different firms is argued to be one of the main carriers of knowledge between firms in a cluster. Chapter 3 empirically examines the role of informal contacts in a specific cluster. In a recent questionnaire, a sample of engineers in the regional cluster of wireless communication firms in Northern Denmark was asked a series of questions on informal networks. It is analyzed whether the engineers actually acquire valuable knowledge through these networks. It is found that the engineers do share even rather complex technical knowledge with informal contacts. This shows that informal contacts are – at least for some engineers – important channels of knowledge diffusion.

In Chapter 4, we go into a deeper study of the theories of knowledge and organizational inheritance as an alternative to the dominant theories of clustering and agglomeration. Recent empirical studies have focused on how capabilities of new entering firms are important for the evolution of industries over time. The performance of new entrants appears to be significantly influenced by their pre-entry background. The general impression of the literature is that firms founded by former employees of successful incumbents have shown higher propensities to survive than other categories of new entrants. In Chapter 4, this approach is used to study the emergence and growth over the past three decades of the wireless telecommunications cluster around Aalborg in North Jutland, Denmark (NorCOM). The aim is to analyze the dominating forces behind the growth of NorCOM using detailed information about the founding events and organizational background of every individual entrant in the cluster. It is found that the technological successes of early firms in the region have powered a spin off process, which can account for the majority of the growth in number of firms and employment in the cluster.

The third part of the thesis studies the aspects of geographical clustering and knowledge diffusion. Chapter 5 is an explorative study of labor mobility between firms in Denmark. Based on the theoretical assumption that knowledge is embodied in people and the movement of people facilitates knowledge flows between firms, we study the extent and mobility of full time employees in six industries and 20 functional urban regions or local labor markets. Using unique data on the total Danish labor market in the period from 1990 to 2000, it is possible to track each employee through time. Different types of mobility with respect to regional concentration of industries, education, and occupation are investigated in order to see whether concentrations of industries in particular regions have higher mobility rates and explore where these concentrations mainly hire new employees from. In the final part of the chapter a logistic model is presented in order to predict intra-industry mobility across regions on the basis of several sociological variables.

In Chapter 6, we look at specific patterns and dependencies of firm growth in relation to industry and geographical location of the individual firm. It is investigated whether there are any underlying relationships concerning firm growth rates or whether they should be viewed as part of a random process. This is relevant for studying firm performance and knowledge diffusion in the future. Using Danish firm data covering almost 9,000 observations, key variables, such as size, age, and industry structure are tested against firm growth rates in turnover and employment in order to find a valid relationship. Besides running the regressions on all observations, we also consider and find interesting patterns in geographical and industry contexts. Thus, the conclusion is that firm growth cannot be considered entirely idiosyncratic. Firm growth is highly dependent on industry and geography, and this should be considered, when examining the performance of firms relative to the different aspects of knowledge diffusion as argued in several of the above chapters.

The fourth part of the thesis concerns the importance of local universities and the conclusion of the thesis. Here Chapter 7 compares university-industry links in three Nordic peripheral regions with fairly new but already broad and comprehensive universities that all include rather big technical faculties (engineering schools). The regions include the area around Luleå in northern Sweden, Aalborg in northern Denmark, and Oulu in northern Finland. The three regions have somehow similar demographic characteristics. The comparison shows that the technical faculties of the three universities have strong linkages with the local industry, but they are of a different nature. In all three cases the industrial activities in the region are fairly closely connected to the activities of the local universities. The chapter is concluded by a discussion of some common issues concerning how university-industry links work, especially concerning their role in clusters.

Chapter 8 finalizes the thesis with a discussion of the general conclusions and policy perspectives. The research questions are discussed in turn with reference to the general applicability of the findings in a theoretical perspective. This includes a motivation for future research and important challenges for the future. Clusters as economic phenomena are more than ever a desired target of policy makers around the globe as mention above. In this respect, Chapter 8 includes a discussion on the general policy implications of this thesis with a concrete link to the present efforts of the local government in Aalborg and North Jutland.

Summary in Danish

Denne ph.d. afhandling har oversat til dansk titlen, “Vidensspredning og regionale virksomhedsklynger”. Den tager sit udgangspunkt i forskellige studier af den danske IKT industri og denne geografiske dimension. Kapitel 1 beskriver de dominerende teorier, der anvendes inden for det økonomisk geografiske forskningsfelt. En af de mest dominerende antagelser af denne litteratur er, at viden spredes gennem netværksrelationer i virksomhedsklynger. Denne antagelse er den centrale i denne afhandling og genstand for forskellige empiriske tests i afhandlingens mest centrale kapitler 3-5.

Kapitel 2 giver en gennemgående beskrivelse af styrker og svagheder i den danske IKT sektor. Dette fremstår som noget af et paradoks, hvor stærke positioner på brugersiden er på den ene side. Men på den side står Danmark svagt på udbudssiden i forhold til produktion af IKT udstyr og services. Dette paradoks diskuteres i forhold til andre mindre lande som Sverige og Finland, der står relativt stærkere end Danmark. To regionalt funderede styrkepunkter præsenteres, som få lys-punkter i det danske IKT erhvervsliv.

Afhandlingens centrale Del 2 omhandler kvalitative studier af et af de styrkepunkter, som blev beskrevet i Kapitel 2. Kapitel 3 tester således den helt centrale hypotese om videnspredning fra teorien ved hjælp af et spørgeskema udsendt til alle ingeniører i den nordjyske klynge af virksomheder inden for trådløs kommunikation. Konklusion af dette er, at viden i et vist omfang spredes gennem netværk. Kapitel 4 undersøger en helt anden form for videnspredning, nemlig den spredning, som sker, når en medarbejder forlader en virksomhed for at stifte deres egen i samme branche. Disse nye virksomheder kaldes spinoffs og repræsenterer en spredning af viden indeholdt i personer. Viden som ikke umiddelbart kan overføres ved samtale og nedskrivning. Stifternes baggrund i virksomheder i den nordjyske klynge af trådløse virksomheder kortlægges for at undersøge, hvorvidt spinoffs klarer sig bedre end andre typer af virksomhedsopstarter i kraft af deres viden fra arbejdet i en anden virksomhed i klyngen. Resultatet af denne undersøgelse er, at spinoffs har været en helt central faktor bag udviklingen af klyngen. Spinoffs har ligeledes genereret størstedelen af de arbejdspladser, som findes i klyngen i dag.

Den tredje del af afhandlingen omhander mere kvantitative undersøgelse af klyngedannelse. Kapitel 5 indeholder indledende undersøgelse af mere generel arbejdskraftsmobilitet. Ved hjælp af en unik database indeholdende hele den danske arbejdsstyrke identificeres 39 virksomhedsklynger, hvori mobiliteten undersøges. Resultatet af dette er, at klyngerne ikke viser sig at have en højere mobilitet inden for regionen og industrien, hvilket den dominerende litteratur ellers argumentere for. I stedet viser det sig, at virksomhedsklyngerne generelt hyrer flere nye medarbejdere fra samme industri end regioner uden klynger. Det betyder, at klyngevirksomhederne kan have en fordel ved at kunne hyre flere medarbejdere med industrispecifik viden end andre virksomheder. Et andet resultat er, at arbejdskraften generelt er ret immobil på tværs af regioner. Årsagen til dette søges kortlagt sidst i kapitelet, hvor en logistisk model præsenteres. Modellens estimation viser, at personer generelt er mindre mobile, jo ældre de er, samt jo mere etableret de bliver i forhold til civil status. Dette betyder, at ældre og mere erfaren arbejdskraft ikke så ofte flytter på tværs af regioner.

Kapitel 6 analyser forskellige faktorer bag virksomhedsvækst. En central konklusion i kapitlerne 3-5 er, at netop denne form for analyser bør spille en større rolle i fremtidig forskning i forhold til videnspredning. Kapitlet finder en række helt centrale faktorer, som skal medtages i disse fremtidige analyser.

Den sidste del af afhandlingen bringer en diskussion af universiteters betydning for det lokale erhvervsliv samt bringer konklusionerne og policy implikationerne til afhandlingen. Kapitel 7 sammenligner Luleå (Sverige), Aalborg (Danmark) og Oulu (Finland) for at undersøge universiteterne effekter. Resultatet af denne sammenligning er, at universiteternes vigtigste rolle er af udbyde arbejdskraft.

Kapitel 8 er afhandlingens konklusion, som resumerer resultaterne fra den centrale del af afhandlingen vedrørende tests af den dominerende teori. Derudover diskuteres mulighederne for fremtidig forskning med afsæt i afhandlingen. Slutteligt diskuteres den politiske del af virksomhedsklyngerne.